Belonging & Bots — A Question Worth Asking
What Should AI Never Know About You?
This isn't a security warning. It's a more personal question than that.
Before I tell anyone to start using AI for their work, I ask them one question. Not "what tools have you tried" or "what's your tech comfort level." I ask:
"What do you most want to protect from AI?"
Most people pause. It's not a question they've been asked. And the pause is the point — because the answer tells you more about how to use AI well than any tutorial does.
The protection question isn't about privacy
When I say "protect," I'm not talking about passwords or data security. I'm talking about something more fundamental: the parts of your work that are irreducibly you.
Your judgment about a situation no one else has seen the full context of. The relationship you've built with someone over ten years. The instinct that tells you a project is going wrong before you can articulate why. The way you read a room. The thing you know about your field that you can't fully explain but that makes your work distinctive.
None of that should go to AI. Not because AI is a threat to it — but because handing it off would hollow the work out. It would make you less useful, not more.
What the question reveals
Here's what I've noticed after asking this question hundreds of times: the moment someone can answer it clearly, they stop being afraid of AI.
The fear — the "what if AI takes my job" fear — usually lives in the blur between what you do and who you are. When you name what to protect, the blur clears. You realize how much of your work is genuinely yours and how much of it is just labor that got attached to you because no one else was doing it.
The second thing is: once you know what to protect, the rest becomes obvious. Everything that isn't on that list is something a director can handle.
A few things worth protecting
These aren't universal — your list will be different. But these come up often enough to be worth naming:
Your read on people
AI can summarize a meeting. It can't tell you that someone's enthusiasm felt performative, or that a quiet person in the corner understood more than they let on. That reading is yours.
Your professional instincts
The pattern recognition you've built over years of being in a field. The thing that makes you say "something's off" before you can prove it. AI doesn't have this. You do.
Your relationships
AI can draft a message to someone you haven't talked to in two years. It shouldn't be the one to decide whether to send it, or what the relationship is really worth. That's yours.
Your creative distinctiveness
If your voice is the reason people hire you or follow you or want to work with you — protect it. Use AI to handle the volume, not to replace the voice.
Your ethical judgment
What you're willing to do and not do. What you stand for in your work. AI will do what you ask. That's exactly why the asking has to come from you.
The practical upside
Once you've named what to protect, you have a different problem: a lot of things on your plate that don't require your best self. The writing that needs to get done but doesn't need to be brilliant. The follow-ups that need to be warm but not particularly thoughtful. The operational stuff that just needs to happen.
That's what directors are for. The 1+7 model is essentially a way of acting on the answer to this question — systematically clearing the work that doesn't need to be you, so you can actually be you in the parts that matter.
Start with what's heaviest
The free worksheet helps you identify which part of your work to hand off first — and name the director who handles it.